Executive Compensation as an Agency Problem

Executive Compensation as an AgencyProblem

Lucian Arye Bebchuk and Jesse M. Fried

E xecutive compensation has long attracted a great deal of attention fromfinancial economists. Indeed, the increase in academic papers on thesubject of CEO compensation during the 1990s seems to have outpaced

even the remarkable increase in CEO pay itself during this period (Murphy, 1999).Much research has focused on how executive compensation schemes can helpalleviate the agency problem in publicly traded companies. To understand ade-quately the landscape of executive compensation, however, one must recognizethat the design of compensation arrangements is also partly a product of this sameagency problem.

Alternative Approaches to Executive Compensation

Our focus in this paper is on publicly traded companies without a controllingshareholder. When ownership and management are separated in this way, manag-ers might have substantial power. This recognition goes back, of course, to Berleand Means (1932, p. 139) who observed that top corporate executives, “while inoffice, have almost complete discretion in management.” Since Jensen and Meck-ling (1976), the problem of managerial power and discretion has been analyzed inmodern finance as an “agency problem.”

Managers may use their discretion to benefit themselves personally in a variety

y Lucian Arye Bebchuk is the William J. Friedman Professor of Law, Economics and Finance,Harvard Law School, and Research Associate, National Bureau of Economic Research, bothin Cambridge, Massachusetts. Jesse M. Fried is a Professor of Law at Boalt Hall School ofLaw, University of California at Berkeley, Berkeley, California. Their e-mail addresses are�[email protected]� and �[email protected]�, respectively.

Journal of Economic Perspectives—Volume 17, Number 3—Summer 2003—Pages 71–92

of ways (Shleifer and Vishny, 1997). For example, managers may engage in empirebuilding ( Jensen, 1974; Williamson, 1964). They may fail to distribute excess cashwhen the firm does not have profitable investment opportunities ( Jensen, 1986).Managers also may entrench themselves in their positions, making it difficult tooust them when they perform poorly (Shleifer and Vishny, 1989). Any discussion ofexecutive compensation must proceed against the background of the fundamentalagency problem afflicting management decision-making. There are two differentviews, however, on how the agency problem and executive compensation arelinked.

Among financial economists, the dominant approach to the study of executivecompensation views managers’ pay arrangements as a (partial) remedy to the agencyproblem. Under this approach, which we label the “optimal contracting approach,”boards are assumed to design compensation schemes to provide managers withefficient incentives to maximize shareholder value. Financial economists have donesubstantial work within this optimal contracting model in an effort to understandexecutive compensation practices; recent surveys of this work include Murphy(1999) and Core, Guay and Larcker (2001). To some researchers working withinthe optimal contracting model, the main flaw with existing practices seems to bethat, due to political limitations on how generously executives can be treated,compensation schemes are not sufficiently high-powered ( Jensen and Murphy,1990).

Another approach to studying executive compensation focuses on a differentlink between the agency problem and executive compensation. Under this ap-proach, which we label the “managerial power approach,” executive compensationis viewed not only as a potential instrument for addressing the agency problem butalso as part of the agency problem itself. As a number of researchers have recog-nized, some features of pay arrangements seem to reflect managerial rent-seekingrather than the provision of efficient incentives (for example, Blanchard, Lopez-de-Silanes and Shleifer, 1994; Yermack, 1997; and Bertrand and Mullainathan,2001). We seek to develop a full account of how managerial influence shapes theexecutive compensation landscape in a forthcoming book (Bebchuk and Fried,2004) that builds substantially on a long article written jointly with David Walker(Bebchuk, Fried and Walker, 2002).

Drawing on this work, we argue below that managerial power and rent extrac-tion are likely to have an important influence on the design of compensationarrangements. Indeed, the managerial power approach can shed light on manysignificant features of the executive compensation landscape that have long beenseen as puzzling by researchers working within the optimal contracting model. Wealso explain that managers’ influence over their own pay might impose substantialcosts on shareholders—beyond the excess pay executives receive—by diluting and distorting managers’ incentives and thereby hurting corporateperformance.

Although the managerial power approach is conceptually quite different fromthe optimal contracting approach, we do not propose the former as a complete

72 Journal of Economic Perspectives

replacement for the latter. Compensation arrangements are likely to be shapedboth by market forces that push toward value-maximizing outcomes, and by man-agerial influence, which leads to departures from these outcomes in directionsfavorable to managers. The managerial power approach simply claims that thesedepartures are substantial and that optimal contracting alone cannot adequatelyexplain compensation practices.

The Limitations of Optimal Contracting

The optimal contracting view recognizes that managers suffer from an agencyproblem and do not automatically seek to maximize shareholder value. Thus,providing managers with adequate incentives is important. Under the optimalcontracting view, the board, working in shareholders’ interest, attempts to providecost-effectively such incentives to managers through their compensation packages.

Optimal compensation contracts could result either from effective arm’slength bargaining between the board and the executives or from market con-straints that induce these parties to adopt such contracts even in the absence ofarm’s length bargaining. However, neither of these forces can be expected toprevent significant departures from arm’s length outcomes.1

Just as there is no reason to presume that managers automatically seek tomaximize shareholder value, there is no reason to expect a priori that directors willeither. Indeed, directors’ behavior is also subject to an agency problem, which inturn undermines their ability to address effectively the agency problems in therelationship between managers and shareholders.

Directors generally wish to be re-appointed to the board. Average directorcompensation in the 200 largest U.S. corporations was $152,626 in 2001 (PearlMeyers and Partners, 2002). In the notorious Enron case, the directors were eachpaid $380,000 annually (Abelson, 2001). Besides an attractive salary, a directorshipis also likely to provide prestige and valuable business and social connections. CEOsplay an important role in renominating directors to the board. Thus, directorsusually have an incentive to favor the CEO.

To be sure, in a world in which shareholders selected individual directors,directors might have an incentive to develop reputations as shareholder-serving.However, board elections are by slate, dissidents putting forward their own directorslate confront substantial impediments, and such challenges are therefore exceed-ingly rare (Bebchuk and Kahan, 1990). Typically, the director slate proposed bymanagement is the only one offered.

The key to a board position is thus being placed on the company’s slate.

1 Shareholders could try to challenge undesirable pay arrangements in court. However, corporate lawrules effectively prevent courts from reviewing compensation decisions (Bebchuk, Fried and Walker,2002, pp. 779–781).

Lucian Arye Bebchuk and Jesse M. Fried 73

Because the CEO’s influence over the board gives her significant influence over thenomination process, directors have an incentive to “go along” with the CEO’s payarrangement, a matter dear to the CEO’s heart—at least as long as the compensa-tion package remains within the range of what can plausibly be defended andjustified. In addition, because being on the company’s slate is the key to beingappointed, developing a reputation for haggling with the CEO over compensationwould hurt rather than help a director’s chances of being invited to join othercompanies’ boards. Yet another reason to favor the CEO is that the CEO can affectdirectors’ compensation and perks.

Directors typically have only nominal equity interests in the firm (Baker,Jensen and Murphy, 1988; Core, Holthausen and Larcker, 1999). Thus, even adirector who did not place much value on a board seat would still have littlepersonal motivation to fight the CEO and her friends on the board on compensa-tion matters. Moreover, directors usually lack easy access to independent informa-tion and advice on compensation practices necessary to effectively challenge theCEO’s pay.

Finally, market forces are not sufficiently strong and fine-tuned to assureoptimal contracting outcomes. Markets—including the market for corporate con-trol, the market for capital and the labor market for executives—impose someconstraints on what directors will agree to and what managers will ask them toapprove. An analysis of these markets, however, indicates that the constraints theyimpose are far from tight and permit substantial deviations from optimal contract-ing (Bebchuk, Fried and Walker, 2002).

Consider, for example, the market for corporate control—the threat of atakeover. Firms frequently have substantial defenses against takeovers. For exam-ple, a majority of companies have a staggered board, which prevents a hostileacquirer from gaining control before two annual elections pass, and often enablesincumbent managers to block hostile bids that are attractive to shareholders. Toovercome incumbent opposition, a hostile bidder must be prepared to pay asubstantial premium; during the second half of the 1990s, the average premium inhostile acquisitions was 40 percent (Bebchuk, Coates and Subramanian, 2002). Thedisciplinary force of the market for corporate control is further weakened by theprevalence of “golden parachute” provisions, as well as acquisition-related benefitsthat target managers often receive when an acquisition takes place. The market forcorporate control thus leaves managers with considerable slack and ability toextract private benefits.

To be sure, the market for control might impose some costs on managers whoare especially aggressive in extracting rents; we later note evidence that CEOs offirms with stronger takeover protection get pay packages that are both larger andless sensitive to performance. The important point is that the market for corporatecontrol fails to impose tight constraints on executive compensation.

Some responses to our earlier work assumed that our analysis of the absenceof arm’s length bargaining did not apply to cases in which boards negotiate pay witha CEO candidate from outside the firm (for example, Murphy, 2002). However,


while such negotiations might be closer to the arm’s length model than negotia-tions with an incumbent CEO, they still fall quite short of this benchmark. Amongother things, directors negotiating with an outside CEO candidate know that afterthe candidate becomes CEO, she will have influence over their renomination to theboard and over their compensation and perks. The directors will also wish to havegood personal and working relationships with the person who is expected tobecome the firm’s leader and a fellow board member. And while agreeing to a paypackage that favors the outside CEO hire imposes little financial cost on thedirectors, any breakdown in the hiring negotiations, which might embarrass thedirectors and in any event force them to reopen the CEO selection process, wouldbe personally costly to them. Finally, directors’ limited time forces them to rely oninformation shaped and presented by the company’s human resources staff andcompensation consultants, all of whom have incentives to please the incomingCEO.

The Managerial Power Approach

The very reasons for questioning the ability of optimal contracting toexplain compensation practices adequately also suggest that executives havesubstantial influence over their own pay. In addition, these reasons suggest thatthe greater is managers’ power, the greater is their ability to extract rents. Thereare limits to what directors will accept and what markets will permit, but theseconstraints do not prevent managers from obtaining arrangements that aresubstantially more favorable than those they could obtain by bargaining at arm’slength.

One important building block of the managerial power approach is “outrage”costs and constraints. The tightness of the constraints managers and directorsconfront depends, in part, on how much “outrage” a proposed arrangement isexpected to generate among relevant outsiders. Outrage might cause embarrass-ment or reputational harm to directors and managers, and it might reduce share-holders’ willingness to support incumbents in proxy contests or takeover bids. Themore outrage a compensation arrangement is expected to generate, the morereluctant directors will be to approve the arrangement and the more hesitantmanagers will be to propose it in the first instance. Thus, whether a compensationarrangement that is favorable to executives but suboptimal for shareholders isadopted will depend on how it is perceived by outsiders.

There is evidence that the design of compensation arrangements is indeedinfluenced by how outsiders perceive them. Johnson, Porter and Shackell (1997)find that CEOs of firms receiving negative media coverage of their compensationarrangements during 1992–1994 subsequently received relatively small pay in-creases and had the pay-performance sensitivity of their compensation arrange-ments increased. Thomas and Martin (1999) find that, during the 1990s, CEOs offirms that were the target of shareholder resolutions criticizing executive pay had

Executive Compensation as an Agency Problem 75

their annual compensation reduced over the following two years by an average of$2.7 million.

The potential significance of outsiders’ perception of a CEO’s compensationand of outrage costs explains the importance of yet another building block of themanagerial power approach—“camouflage.” To avoid or minimize the outrage thatresults from outsiders’ recognition of rent extraction, managers have a substantialincentive to obscure and try to legitimize—or, more generally, to camouflage—theirextraction of rents. The strong desire to camouflage might lead to the adoption ofinefficient compensation structures that hurt managerial incentives and firm per-formance. This concept of camouflage turns out to be quite useful in explainingmany otherwise puzzling features of the executive compensation landscape.

The importance of how compensation arrangements are perceived meansthat, in the executive compensation area, the transparency of disclosure mat-ters. Financial economists often focus on the role of disclosure in gettinginformation incorporated into market pricing. It is widely believed that infor-mation can become reflected in stock prices as long as it is known and fullyunderstood by a limited number of market professionals. In the executivecompensation context, however, the ability of plan designers to choose arrange-ments that favor managers depends on how these arrangements are perceivedby a much wider group of outsiders. As a result, the transparency and salienceof disclosure can have a significant effect on CEO compensation.

Murphy (2002) and Hall and Murphy (this issue) argue that our approachcannot explain increases in managerial pay during the 1990s. In their view, CEOpower declined during this period. Given the strengthening of takeover defensesduring the 1990s, however, it is unclear whether CEO power diminished during thisperiod. In any event, executive pay increases during the 1990s resulted not fromchanges in managerial power but rather from other factors, none of which isinconsistent with the managerial power approach.

First, seeking to make pay more sensitive to performance, regulators andshareholders encouraged the use of equity-based compensation. Taking advan-tage of this enthusiasm, executives used their influence to obtain substantialoption pay without giving up corresponding amounts of their cash compensa-tion. Furthermore, the options they received did not link pay tightly to themanagers’ own performance, but rather enabled managers to reap windfallsfrom that part of the stock price increase that was due solely to market andsector trends beyond their control. As a result, managers were able to capturemuch larger gains than more cost-effective and efficient option plans wouldhave provided. Second, because executive compensation has historically beencorrelated with market capitalization, the rising stock markets of the 1990s,which carried along with them even many poorly performing companies, pro-vided a convenient justification at most firms for substantial pay increases.Third, market booms weaken outrage constraints; exuberant shareholders areless likely to scrutinize and resent generous pay arrangements, in the same waythat the recent market declines have made shareholders more prone to do so.


Power and Camouflage at Work

We illustrate below the potential value of the managerial power approach bydiscussing four patterns and practices that can be at least partly explained by powerand camouflage: the relationship between power and pay; the use of compensa-tion consultants; stealth compensation; and gratuitous goodbye payments to de-parting executives.

Power-Pay RelationshipsThe managerial power approach predicts that pay will be higher and/or less

sensitive to performance in firms in which managers have relatively more power.Other things being equal, managers would tend to have more power when i) theboard is relatively weak or ineffectual; ii) there is no large outside shareholder;iii) there are fewer institutional shareholders; or iv) managers are protected byantitakeover arrangements. There is evidence indicating that each of these factorsaffects pay arrangements in the way predicted by the managerial power approach.

Executive compensation is higher when the board is relatively weak or ineffectualvis-a-vis the CEO. Core, Holthausen and Larcker (1999) find that CEO compensa-tion is higher under the following conditions: when the board is large, which makesit more difficult for directors to organize in opposition to the CEO; when more ofthe outside directors have been appointed by the CEO, which could cause them tofeel a sense of gratitude or obligation to the CEO; and when outside directors serveon three or more boards, and thus are more likely to be distracted. Also, CEO payis 20–40 percent higher if the CEO is the chairman of the board (Cyert, Kang andKumar, 2002; Core, Holthausen and Larcker, 1999). Finally, CEO pay is negativelyrelated to the share ownership of the board’s compensation committee; doublingcompensation committee ownership reduces nonsalary compensation by 4–5 per-cent (Cyert, Kang and Kumar, 2002).

The presence of a large outside shareholder is likely to result in closer monitoring(Shleifer and Vishny, 1986), and it can be expected to reduce top managers’influence over their compensation. Consistent with this observation, Cyert, Kangand Kumar (2002) find a negative correlation between the equity ownership of thelargest shareholder and the amount of CEO compensation: doubling the percent-age ownership of the outside shareholder reduces nonsalary compensation by12–14 percent. Bertrand and Mullainathan (2000) find that CEOs in firms that lacka 5 percent (or larger) external shareholder tend to receive more “luck-based”pay—pay associated with profit increases that are entirely generated by externalfactors (such as changes in oil prices and exchange rates) rather than by managers’efforts. They also find that in firms lacking large external shareholders, the cashcompensation of CEOs is reduced less when their option-based compensation isincreased. Relatedly, in an examination of Standard & Poor’s 500 firms during theperiod 1992–1997, Benz, Kucher and Stutzer (2001) find that a higher concentra-tion of shareholders results in a significantly smaller amount of options grants totop executives.


A larger concentration of institutional shareholders might result in greater monitor-ing and scrutiny of the CEO and the board. Examining CEO pay in almost 2000firms during the period 1991–1997, Hartzell and Starks (2002) find that the moreconcentrated is institutional ownership, the lower is executive compensation. Theyalso find that a larger institutional presence results in more performance-sensitivecompensation. Examining CEO compensation in the 200 largest companies during1990–1994, David, Kochar and Levitas (1998) find that the effect of institutionalshareholders on CEO pay depends on the types of relationships they have with thefirm. They divide institutional shareholders into 1) those that have no otherbusiness relationship with the firm and are thus concerned only with the firm’sshare value (“pressure-resistant” institutions); and 2) those that have other businessrelationships with the firm (like managing a pension fund) and are thus vulnerableto management pressure (“pressure-sensitive” institutions). As the managerialpower approach predicts, CEO pay is negatively correlated with the presence ofpressure-resistant institutional investors and positively correlated with the presenceof pressure-sensitive ones.

The adoption of antitakeover provisions makes CEOs less vulnerable to a hostiletakeover. Borokhovich, Brunarski and Parrino (1997), examining 129 firms thatadopted antitakeover provisions (such as a supermajority rule) during the period1979–1987, find that CEOs of firms adopting such provisions enjoy above-marketcompensation before adoption of the antitakeover provisions and that adoption ofthese provisions increases their excess compensation significantly. This pattern isnot readily explainable by optimal contracting; indeed, if managers’ jobs are moresecure, shareholders should be able to pay managers a lower risk premium(Agrawal and Knoeber, 1998). In another study, Cheng, Nagar and Rajan (2001)find that CEOs of Forbes 500 firms that became protected by state antitakeoverlegislation enacted during the period 1984–1991 reduced their holdings of sharesby an average of 15 percent, apparently because the shares were not as necessary formaintaining control. Optimal contracting might predict that a CEO protected byantitakeover legislation would be required to buy more shares to restore the CEO’sincentive to increase shareholder value.

Compensation ConsultantsU.S. public companies typically employ outside consultants to provide input

into the executive compensation process (Bizjack, Lemmon and Naveen, 2000).The use of consultants can be explained within the optimal contracting frameworkon grounds that they supply useful information and contribute expertise on thedesign of compensation packages. But although compensation consultants mightplay a useful role, they also can help in camouflaging rents. The incentives ofcompensation consultants—and the evidence regarding their use—suggest thatthese consultants are often used to justify executive pay rather than to optimize it.

Compensation consultants have strong incentives to use their discretion tobenefit the CEO. Even if the CEO is not formally involved in the selection of thecompensation consultant, the consultant is usually hired by the firm’s human


resources department, which is subordinate to the CEO. Providing advice that hurtsthe CEO’s pocketbook is hardly a way to enhance the consultant’s chances of beinghired in the future by this firm or, indeed, by any other firms. Moreover, executivepay specialists often work for consulting firms that have other, larger assignmentswith the hiring company, which further distorts their incentives (Crystal, 1991).

Pay consultants can favor the CEO by providing the compensation data thatare most useful for justifying a high level of pay. For example, when firms do well,consultants argue that pay should reflect performance and should be higher thanthe average in the industry—and certainly higher than that of CEOs who are doingpoorly. In contrast, when firms do poorly, the consultants focus not on perfor-mance data but rather on peer group pay to argue that CEO compensation shouldbe higher to reflect prevailing industry levels (Gillan, 2001).

After the compensation consultant has collected and presented the “relevant”comparative data, the board generally sets pay equal to or higher than the medianCEO pay in the comparison group. Reviewing the reports of compensation com-mittees in 100 large companies, Bizjack, Lemmon and Naveen (2000) report that96 used peer groups in determining management compensation and that a largemajority of firms that use peer groups set compensation at or above the fiftiethpercentile of the peer group. The combination of helpful compensation consult-ants and sympathetic boards is partly responsible for the widely recognized “ratch-eting up” of executive salaries (Murphy, 1999, p. 2525).

After the board approves the compensation package, firms use compensationconsultants and their reports to justify executive compensation to shareholders.Examining Standard & Poor’s 500 companies during the period 1987–1992, Wade,Porac and Pollack (1997) find that companies that pay their CEOs larger basesalaries, and firms with more concentrated and active outside ownership, aremore likely to cite the use of surveys and consultants in justifying executive payin their proxy reports to shareholders. This study also finds that, when account-ing returns are high, firms emphasize the accounting returns and downplaymarket returns.

Stealth CompensationAs we document in Bebchuk and Fried (2003), firms use pay practices that

make less transparent the total amount of executive compensation and the extentto which compensation is decoupled from managers’ own performance. Amongthe arrangements used by firms that camouflage the amount and the performance-insensitivity of compensation are pension plans, deferred compensation, post-retirement perks, and consulting contracts.

Most of the pension and deferred compensation benefits given to executivesdo not enjoy the large tax subsidy that applies to the standard retirement arrange-ments provided to other employees. In the case of executives, such arrangementslargely shift tax liability from the executive to the firm in ways that sometimes evenincrease the joint tax liability of the two parties. The efficiency grounds forproviding compensation through in-kind retirement perks and guaranteed postre-


tirement consulting fees are also far from clear. All of these arrangements, however,make pay less salient.

Among other things, under existing disclosure rules, firms do not have to placea dollar value on—and include in the firm’s publicly filed compensation tables—compensation provided to executives after they retire. Although the existence ofexecutives’ retirement arrangements must be noted in certain places in the firm’spublic filings, this disclosure is less salient, because outsiders focus on the dollaramounts reported in the compensation tables. Indeed, the compensation tablenumbers are used by the ExecuComp database, which is the basis for much of theempirical work on executive compensation.

Another practice with camouflage benefits was the use of executive loans.While the Sarbanes-Oxley Act of 2002 now prohibits such loans, prior to the Act’sadoption more than 75 percent of the 1,500 largest U.S. firms lent money toexecutives (King, 2002). It is not readily apparent that having firms (rather thanbanks) lend to executives—or that providing compensation in the form of favor-able interest rates—is efficient. But loans are useful for reducing the salience ofmanagers’ compensation.

To begin with, the implicit compensation provided by below-market-rate loansoften does not appear in the compensation tables in the firm’s annual filing. TheSEC ruled that firms must disclose in the category of “other annual compensation”the difference between the interest actually paid on executive loans and the“market rate.” However, the SEC did not define “market rate,” and firms haveinterpreted the term in a manner that enabled them to exclude the value of largeinterest rate subsidies from the compensation tables.

For example, WorldCom did not report in its compensation tables any incometo CEO Bernard Ebbers from the over $400 million of loans he received fromWorldCom at an interest rate of 2.15 percent; it later justified the omission on thegrounds that 2.15 percent was the “market rate” at which WorldCom was borrowingunder one of its credit facilities. However, 2.15 percent was far below the more than5 percent rate that Ebbers would have paid at that time in the market to borrowfunds. To be sure, the existence and terms of the loans (although not an estimateof the conferred benefits) had to be noted elsewhere in the firm’s public filings asa related party transaction. However, this disclosure is much less salient becauseoutsiders interested in executives’ compensation commonly focus on the compen-sation tables. Indeed, in Ebbers’s case, despite the large financial benefit providedby the extremely low interest rate on his loans, the loans received no mediaattention and no outside scrutiny until WorldCom became involved in an account-ing scandal.

Another manner in which loans provided camouflage was through the practiceof loan forgiveness. A firm that gave an executive a loan to buy a large amount ofstock would often not demand full repayment of the loan if the stock value fellbelow the amount due on the loan. As a result, the arrangement was similar to (but,it can be shown, often less tax efficient than) granting the executive an option tobuy shares at a price equal to the amount owed on the loan. However, option grants


must be reported the year they are made in the firm’s publicly filed compensationtables. In contrast, when granting a loan that likely will be forgiven if the stock pricedrops, the firm did not need to include the option value of the arrangement in thecompensation tables in the year the loan was made. Indeed, if the stock price fell,the loan would often be forgiven at the time that the executive left the company,when any resulting outrage is likely to have little impact on the executive person-ally. For example, George Shaheen, the Webvan CEO who resigned shortly beforeWebvan went bankrupt, had a $6.7 million loan forgiven in exchange for $150,000of Webvan stock (Lublin, 2002).

Gratuitous Goodbye PaymentsIn many cases, boards give departing CEOs payments and benefits that are

gratuitous—not required under the terms of the CEO’s compensation contract.Such gratuitous goodbye payments are common even when CEOs perform sopoorly that the board feels compelled to replace them.

Compensation contracts usually provide executives with generous severancearrangements even when they depart following very dismal performance. Such “softlanding” provisions provide executives with insurance against being fired due topoor performance. It is far from clear that these arrangements reflect optimalcontracting; after all, such provisions reduce the difference in managerial payoffsbetween good and poor performance that firms spend so much money trying tocreate. Our focus here, however, is on payments that go beyond the severancearrangements that are contractually specified.

For example, when Mattel CEO Jill Barad resigned under fire, the boardforgave a $4.2 million loan, gave her an additional $3.3 million in cash to coverthe taxes for forgiveness of another loan and allowed her unvested options tovest automatically. These gratuitous benefits were in addition to the consider-able benefits that she received under her employment agreement, which in-cluded a termination payment of $26.4 million and a stream of retirementbenefits exceeding $700,000 per year.

It is not easy to reconcile such gratuitous payments with the arm’s length,optimal contracting model. The board has the authority to fire the CEO and paythe CEO her contractual severance benefits. Thus, there is no need to “bribe” apoorly performing CEO to step down. In addition, the signal sent by the goodbyepayment will, if anything, only weaken the incentive of the next CEO to perform.

The making of such gratuitous payments, however, is quite consistent with theexistence of managerial influence over the board. Because of their relationshipwith the CEO, some directors might be unwilling to replace the CEO unless she istreated very generously. Other directors might be willing to replace the CEO in anyevent, but prefer to accompany the move with a goodbye payment to reduce thediscomfort they feel firing the CEO or to make the difficult separation processmore pleasant and less contentious. In all of these cases, directors’ willingness tomake gratuitous payments to the (poorly performing) CEO results from the CEO’srelationship with the directors.


It is important to note that, taking managerial power as given, providinggratuitous payments to fired CEOs might be beneficial to shareholders in someinstances. If many directors are loyal to the CEO, such payments might be necessaryto assemble a board majority in favor of replacing him. In such a case, the practicewould help shareholders when the CEO’s departure is more beneficial to share-holders than the cost to them of the goodbye payment. For our purposes, however,what is important is that these gratuitous payments, whether they are beneficial toshareholders or not, reflect the existence and significance of managerial influence.

Suboptimal Pay Structures

Pay Without PerformanceOptimal contracting arrangements might call for very large amounts of com-

pensation for executives, if such compensation is needed to provide managers withpowerful incentives to enhance shareholder value (Jensen and Murphy, 1990). Theproblem with current arrangements, however, is that the generous compensationprovided executives is linked only weakly to managerial performance. This pay-performance disconnect is puzzling from an optimal contracting view.

The substantial part of compensation that is not equity based has long beencriticized as weakly linked to managerial performance. During the 1990s, therewas no significant correlation between a CEO’s salary and bonus and her firm’sindustry-adjusted performance (Murphy, 1999). In addition, there is evidencethat cash compensation increases when firm profits rise for reasons that clearlyhave nothing to do with managers’ efforts (Blanchard, Lopez-de-Silanes andShleifer, 1994; Bertrand and Mullainathan, 2001). Furthermore, managersreceive substantial non-equity compensation through arrangements that havereceived little attention from financial economists—such as pensions, deferredpay and loans—and this compensation is also relatively insensitive to managers’own performance.

In light of the historically weak link between managers’ performance andtheir non-equity compensation, shareholders and regulators have increasinglylooked to equity-based compensation to provide the desired link between payand performance. In the early 1990s, institutional investors and federal regu-lators sought to encourage the use of such compensation, and the last decadehas witnessed a dramatic growth in the use of stock options. Unfortunately,however, managers have been able to use their influence to obtain option plansthat appear to deviate substantially from optimal contracting in ways that favormanagers.

We wish to emphasize our strong support for the concept of equity-basedcompensation which, if well designed, could provide managers with very desirableincentives. The devil, however, is in the details. Below, we discuss several importantfeatures of existing option compensation plans that are difficult to justify from anoptimal contracting perspective, but can readily be explained by the managerial


power approach: the failure of option plans to filter out windfalls, the almostuniform use of at-the-money options and the broad freedom given to managers tounload options and shares.

It might be asked why risk-averse managers would not use their influence to gethigher cash salaries rather than options. Holding the expected value of additionalcompensation constant, managers would indeed prefer to take the cash. Butmanagers seeking to increase their pay during the 1990s did not have a choicebetween additional compensation in the form of cash and additional compensationin the form of options with the same expected value. Rather, outsiders’ enthusiasmabout equity-based compensation enabled managers to obtain additional compen-sation in the form of options without offsetting reduction in cash compensation.Furthermore, the possible benefits from improved incentives provided defensiblereasons for very large amounts of additional compensation. While Apple CEO SteveJobs was recently able to obtain an option package worth over half a billion dollars,albeit with some outcry, cash compensation of this order of magnitude is (still)quite inconceivable. The fact that better designed options could have providedmuch more cheaply the same incentives has not been sufficiently salient to makeconventional plans patently unjustifiable.

Option Plans that Fail to Filter Out “Windfalls”One widespread and persistent feature of stock option plans is that they fail to

filter out stock price rises that are due to industry and general market trends andthus completely unrelated to managers’ own performance. Under conventionaloption plans, when the market or sector rises substantially, even executives whosecompanies perform poorly relative to the market or sector average can make largeprofits. Paying managers substantial compensation for stock price increases thathave nothing to do with their own performance is difficult to explain under optimalcontracting. The substantial amount currently spent on rewarding managers formarket or sector rises could either be used to enhance incentives (for example, bygiving managers a large number of options linked more tightly to managers’ ownperformance) or be saved with little weakening of incentives.

There are many different ways of designing what we call “reduced-windfall”option plans—plans that filter out all or some of the part of the stock price increasethat is unrelated to managers’ own performance. One approach discussed fre-quently by academics is linking the exercise price of options to a market-wide indexor a sector index (for example, Rappaport, 1999). Another strategy is to conditionthe “vesting” of options on the firm meeting specified performance targets. Thesetargets can be linked to the stock price, earnings per share or any other measure offirm performance.

When the exercise price of an indexed option is linked to market or sectoraverages, there is a substantial possibility that the manager will receive no payofffrom the option plan. If this possibility were regarded as undesirable, reduced-windfall options could easily be designed to produce a high likelihood ofpayout. For example, the exercise price could be indexed not to changes in the


industry or market average, but rather to changes in a somewhat lower bench-mark—say, the stock price of the firm that is at the bottom 20th percentile of theindustry or market. Under such an option plan, executives would have, onaverage, an 80 percent probability of outperforming the benchmark and receiv-ing a payout. But executives would not profit, as they could under conventionalplans, when their performance places them at or below the 20th percentile.

Given the wide variety of reduced-windfall options available and their potentialbenefits, it would probably be optimal in many firms to filter out at least some of theincrease in the stock price that has nothing to do with managers’ own performance.Yet almost all U.S. firms use conventional stock options under which managerscapture the full increase in stock price. In 2001, only about 5 percent of the 250largest U.S. public firms used some form of reduced-windfall options (Levinsohn,2001).

Financial economists have made substantial efforts to develop optimal-contracting explanations for why firms do not use reduced-windfall options. Wesurvey the various explanations in our earlier work (Bebchuk, Fried and Walker,2002, pp. 803–809) and conclude that none of them can adequately explain thewidespread failure to screen out windfalls. From the perspective of managerialpower, however, the failure to filter out general market or industry effects is not atall puzzling. Under this approach, compensation schemes are designed to benefitexecutives without being perceived as clearly unreasonable. Given that using con-ventional options will be legitimate and acceptable (after all, most firms use them)and that moving to indexing or any other form of reduced-windfall options is likelyto be costly or inconvenient for managers, the lack of any real movement towardsuch options is consistent with the managerial power approach.

At-the-Money OptionsAlmost all of the stock options used to compensate executives are “at-the-

money”—that is, their exercise price is set to the grant-date market price (Murphy,1999, p. 2509). An optimally designed scheme would seek to provide risk aversemanagers with cost-effective incentives to exert effort and make value-maximizingdecisions. The optimal exercise price under such a scheme would depend on amultitude of factors that are likely to vary from executive to executive, fromcompany to company, from industry to industry and from time to time. Such factorsmight include the degree of managerial risk aversion (which in turn might beaffected by the manager’s age and wealth), the project choices available to thecompany, the volatility of the company’s stock, the expected rate of inflation andthe length of the manager’s contract, among other things. There is no reason toexpect that “one size fits all”—that the same exercise price is optimal for allexecutives at all firms, in all industries and at all times.

It is therefore highly unlikely that out-of-the-money options—options whoseexercise price is above the current market price—are never optimal. Out-of-the-money options have a lower expected value than at-the-money options because they


are less likely to pay off than at-the-money options, and when they do pay off theholder receives less value. Thus, for every dollar of expected value, a firm can givean executive more out-of-the money options than at-the-money options. By givingmore out-of-the money options, the firm can increase the reward to the managerfor doing particularly well. Out-of-the-money options thus can offer much higherpay-for-performance sensitivity per dollar of expected value than conventionaloptions (Hall, 1999). There is even evidence suggesting that giving managersout-of-the-money options rather than at-the-money-options would, on average,boost firm value (Habib and Ljungqvist, 2000). The almost uniform use of at-the-money options is thus difficult to explain from an optimal contracting perspective.Indeed, economists working within optimal contracting have called this practice a“puzzle” (Hall, 1999, p. 43).

The near uniform use of at-the-money options is not puzzling, however, whenexamined under the managerial power approach. Everything else equal, executivesprefer a lower exercise price. Because at-the-money options might sometimes beoptimal and are employed by almost all other firms, their use in any given case willnot generate outrage. Therefore, compensation plan designers have little reason toincrease the exercise price above the grant-date market price.

Executives would be even better off, of course, if stock options were issued withan exercise price below the grant-date market price. However, such in-the-moneyoptions would create a salient windfall and might generate some outrage costs.Furthermore, in-the-money options would trigger a charge to accounting earnings,which might undermine a main excuse for not using indexed options or otherreduced-windfall options—that the use of such options would hurt reported earn-ings (Bebchuk, Fried and Walker, 2002). Because in-the-money options would thusbe difficult or costly for plan designers to use, and at-the-money options are themost favorable to managers within the remaining range of possibilities, a uniformuse of at-the-money options is consistent with the managerial power approach.

Managers’ Freedom to Unwind Equity IncentivesAnother problem for the optimal contracting approach is managers’ broad

freedom to unload their options and shares. When managers unwind their equityincentives, restoring pay-performance sensitivity requires giving them new optionsor shares. Thus, such unwinding either (1) weakens managers’ incentives or (2)forces the firm to give managers new equity incentives to restore incentives to thepre-unwinding level.

Although an executive becomes entitled to options once they have vested, thecompensation contract could preclude the executive from “cashing out” the vestedoptions—that is, from exercising the options and then selling the acquired shares—for a specified period after the vesting date. Such a limitation would maintainincentives for that additional period without requiring the firm to grant newoptions to replace the ones cashed out.

To be sure, restrictions on executives’ ability to cash out vested equity instru-ments impose liquidity and diversification costs that must be balanced against the


incentive benefits of restricting unwinding. The efficient arrangement is thus likelyto vary from case to case, depending on the executive’s and the firm’s character-istics. But there is no reason to expect that optimal contracts would generally makethe vesting date and the cash-out date identical.

In practice, however, the date on which options vest and the date on whichthey are exercisable are almost always the same. A minority of firms have created“target ownership plans” that require managers to hold a certain amount ofshares (Core and Larcker, 2002). But the targets tend to be rather low, andthere often appears to be no penalty imposed for missing them. As a result ofweak restrictions on unwinding, managers exercise many of their options wellbefore the options expire and sell almost all of the shares thereby acquired(Carpenter, 1998; Ofek and Yermack, 2000). Shares that are not sold after optionexercise are often hedged or partially hedged in transactions that are notreported to the Securities and Exchange Commission (Bettis, Bizjack and Lemmon,2001).

Managers also typically have freedom to determine the precise time of un-winding, a practice that is also difficult to explain under optimal contracting.Although trading on “material” inside information is illegal, the definition of“materiality” and the difficulties of enforcement are such that managers makingselling decisions can use their superior knowledge about the firm with little fear ofliability (Fried, 1998). As a result, managers are able to obtain abnormal returnstrading in their firm’s shares (Seyhun, 1998). It is far from clear, however, thatenabling managers to make such profits is an efficient form of compensation.

Even assuming it were desirable to permit managers to unload shares at acertain stage in their contracts, it does not follow that executives should haveabsolute control over the exact timing of their sales. After all, liquidity or diversi-fication needs are unlikely to arise unexpectedly one morning. Firms could requirethat sales be carried out gradually over a specified period, perhaps pursuant to apre-arranged plan. Alternatively, firms could require executives to disclose publiclyin advance their intended trades, which would reduce executives’ ability to profitfrom their informational advantage (Fried, 1998). Yet firms generally do notimpose any such restrictions.

Because a firm can be held liable if it fails to take reasonable steps to preventinsider trading by its employees, a number of firms have adopted “trading windows”and “blackout periods” to restrict the times during the year when a manager cansell or buy shares (Bettis, Coles and Lemmon, 2000). However, many firms have notput such restrictions in place. And even in firms that have imposed such restric-tions, managers who know undisclosed bad news during a trading window may usethat trading opportunity to unwind a substantial amount of their holdings. Thus,executives retain the ability to dump shares before bad news becomes public. Inone notorious case, Enron insiders sold hundreds of millions of shares beforeinformation about Enron’s actual financial condition was released and the stockprice collapsed.

Although managers’ ability to unwind equity incentives early and to control


the time of such unwinding cannot easily be explained under optimal contracting,it is quite consistent with the managerial power approach. Broad freedom tounload equity instruments provides managers with substantial benefits that are notparticularly conspicuous. The corresponding costs to shareholders from dilutedincentives are also not salient. Furthermore, and perhaps most importantly, man-agers’ unwinding of options and shares provides a convenient justification forfrequently granting managers new equity-based incentives, thereby boosting theirtotal compensation. Although a system of constant unwinding and replenishingincentives is more costly to shareholders than one that requires managers to holdoptions and shares for longer periods, it is obviously much better for managers.

The “Perceived Cost” ExplanationMurphy (2002) and Hall and Murphy (this issue) put forward a “perceived

cost” explanation for the use of conventional, at-the-money options. According totheir explanation, executives and directors erroneously perceive conventional op-tions to be “cheap” or even “nearly free to grant” because such options can begranted without any cash outlay and without reducing reported earnings.

We doubt that executives and their advisers cannot grasp the costs of conven-tional options to shareholders. Assuming that Hall and Murphy are correct insuggesting that managers believe that the stock market is influenced by accountingnumbers rather than underlying economic reality, this would at most mean thatexecutives believe that investors underestimate or ignore the costs of options thatare not expensed for accounting purposes—not that executives themselves fail tosee the significant economic costs that conventional options impose on sharehold-ers (whose ownership interest the options dilute).

One might even be skeptical that directors, many of whom are executivesthemselves, fail to understand the costs of options to shareholders. Indeed, ifdirectors had so little financial sophistication, then the board-monitoring model ofcorporate governance is in even worse shape than our analysis suggests. Let ussuppose, however, that directors have been oblivious to the true cost of conven-tional options. If so, such a misperception on the part of directors is best seen notas an alternative to the managerial power explanation, but rather as one of thefactors contributing to managers’ ability to exert considerable influence over theterms of their pay.

As we discussed earlier, there are several reasons why boards cannot beexpected to engage in arm’s length negotiations with the CEO over executivecompensation; one of them is directors’ lack of easy access to accurate, unbiasedinformation. To the extent directors in fact did misperceive the cost of options,such a misperception would simply be part of the informational problem thatcontributes to directors’ willingness to approve suboptimal arrangements. If direc-tors were ignorant about such an important and widely discussed issue as the actualcost of options, they would likely be inadequately informed about other features ofcompensation arrangements.

In our view, inadequate information is only one of the factors, alongside


inadequate incentives and others, that might lead directors to agree to pay arrange-ments that favor managers. For one thing, directors’ confusion over the cost ofoptions cannot explain the systematic relationship between power and pay andmanagers’ efforts to make compensation less salient that we discussed earlier. Formany purposes, however, it does not matter whether directors’ willingness to acceptarrangements that favor executives is the result of conscious favoritism, honestmisperceptions, inadequate incentives to exert effort, or some combination ofthese factors. The important thing is that directors do not adequately representshareholders’ interests in bargaining with managers over their pay and that thesepay arrangements consequently depart from the arm’s length model in directionsfavorable to executives.

Costs to ShareholdersWhat are the costs imposed on shareholders by managers’ influence over

their own pay? To begin with, there is the excess pay managers receive as a resultof their power—the difference between what managers’ influence enables themto obtain and what they would receive under an arm’s length arrangement.Some might think that this problem is only symbolic, and that these rentshave little effect on shareholders’ bottom line. But a close look at theamounts involved indicates that they add up to much more than small change.In 2000, CEO compensation was on average 7.89 percent of corporate profits inthe firms making up the 1500-company ExecuComp dataset (Balsam, 2002,p. 262).

Furthermore, and perhaps more importantly, managers’ ability to influencetheir pay leads to compensation arrangements that generate worse incentivesthan those that arm’s length contracts would provide. Managers have an interestin compensation schemes that camouflage the extent of their rent extractionor that put less pressure on them to reduce slack. As a result, managerialinfluence might lead to the adoption of compensation arrangements that pro-vide weak or even perverse incentives. In our view, the reduction in shareholdervalue caused by these inefficiencies, rather than the excess rents captured bymanagers, could be the largest cost arising from managers’ ability to influence theircompensation.

To begin, compensation arrangements currently provide weaker incentives toreduce managerial slack and increase shareholder value than likely would beprovided by arm’s length arrangements. As explained, both the non-equity andequity components of managers’ compensation are substantially more decoupledfrom managers’ own performance than appearances might suggest. Shareholdersthus might benefit substantially from the improved performance that a movetoward optimal contracting arrangements could generate.

Prevailing practices not only fail to provide cost-effective incentives to reduceslack, but also create perverse incentives. For one thing, they provide managers’incentives to change firm parameters in a way that would justify increases in pay.


Consider, for example, the familiar problem of empire building. It is commonlybelieved that the practice of granting options provides managers with incentivesnot to undertake acquisitions that are value-decreasing for shareholders. This isclearly the case, however, only in a static model in which all option grants aremade before managers make acquisition decisions. In a dynamic model, managersconsidering an expansion decision that is somewhat value decreasing for share-holders would have different incentives: While such an expansion would reducethe value of their current options, it may raise their aggregate future compensationby an even greater amount because a larger firm size can be used to justify higherpay.

Furthermore, managers’ broad freedom to unload equity incentives can pro-duce substantial inefficiencies. Executives who expect to unload their shares oroptions have weaker incentives to exert effort when the payoff is not going to berecognized by the market at the time they unwind their equity positions (Bar-Gilland Bebchuk, 2003a). Such executives also have incentives to misreport corporateperformance and suppress bad news (Bar-Gill and Bebchuk, 2002). Indeed, suchexecutives also have an incentive to choose projects that are less transparent or toreduce the transparency of existing projects (Bar-Gill and Bebchuk, 2003b). Theefficiency costs of such distortions might exceed, possibly by a large margin,whatever liquidity or risk bearing benefits executives obtain from being able tounload their options and shares at will.

Conclusion

There are good theoretical and empirical reasons for concluding that mana-gerial power substantially affects the design of executive compensation in compa-nies with a separation of ownership and control. Executive compensation can thusbe fruitfully analyzed not only as an instrument for addressing the agency problemarising from the separation of ownership and control—but also as part of theagency problem itself.

The conclusion that managerial power and rent extraction play an importantrole in executive compensation has significant implications for corporate gover-nance, which we explore in our forthcoming book (Bebchuk and Fried, 2004). Itis important to note, however, that this is an area in which widespread recognitionof the problem might contribute to alleviating it. The extent to which managerialinfluence can move compensation arrangements away from optimal contractingoutcomes depends on the extent to which market participants, especially institu-tional investors, recognize the problems we have discussed. Financial economistscan thus make an important contribution to improving compensation arrange-ments by analyzing how current practices deviate from those suggested by optimalcontracting. We hope that future studies of executive compensation will devote tothe role of managerial power as much attention as the optimal contracting modelhas received.


y The authors are grateful to Brad De Long, Andrei Shleifer, Timothy Taylor and MichaelWaldman for many valuable suggestions. For financial support, the authors wish to thank theJohn M. Olin Center for Law, Economics, and Business (Bebchuk) and the Boalt Hall Fundand U.C. Berkeley Committee on Research (Fried).

References

Abelson, Reid. 2001. “Enron Board ComesUnder a Storm of Criticism.” New York Times.December 16, C4.

Agrawal, Anup and Charles R. Knoeber. 1998.“Managerial Compensation and the Threat ofTakeover.” Journal of Financial Economics. 47:2,pp. 219–39.

Baker, George P., Michael C. Jensen andKevin J. Murphy. 1988. “Compensation and In-centives: Practice vs. Theory.” Journal of Finance.63:3, pp. 593–616.

Balsam, Steven. 2002. An Introduction to Execu-tive Compensation. San Diego: Academic Press.

Bar-Gill, Oren and Lucian Arye Bebchuk.2002. “Misreporting Corporate Performance.”Working Paper No. 400, Harvard Olin Centerfor Law, Economics, and Business; availableat �http://papers.ssrn.com/sol3/papers.cfm?abstract_id�354141�.

Bar-Gill, Oren and Lucian Arye Bebchuk.2003a. “The Costs of Permitting Managers toSell Shares.” Mimeo, Harvard Law School.

Bar-Gill, Oren and Lucian Arye Bebchuk.2003b. “Corporate Transparency and Manage-rial Incentives.” Mimeo, Harvard Law School.

Bebchuk, Lucian Arye and Jesse M. Fried.2003. “Stealth Compensation.” Mimeo, HarvardLaw School and Boalt Hall School of Law.

Bebchuk, Lucian Arye and Jesse M. Fried.2004. Pay without Performance. Forthcoming.

Bebchuk, Lucian Arye and Marcel Kahan.1990. “A Framework for Analyzing Legal PolicyTowards Proxy Contests.” California Law Review.October, 78, pp. 1071–135.

Bebchuk, Lucian, John Coates IV and GuhanSubramanian. 2002. “The Powerful AntitakeoverForce of Staggered Boards: Theory, Evidence,and Policy.” Stanford Law Review. May, 54:5, pp. 887–951.

Bebchuk, Lucian Arye, Jesse M. Fried and

David I. Walker. 2002. “Managerial Power andRent Extraction in the Design of Executive Com-pensation.” University of Chicago Law Review. 69:3,pp. 751–846.

Benz, Matthias, Marcel Kucher and Alois Stutzer.2001. “Stock Options: The Managers’ Blessing: In-stitutional Restrictions and Executive Compensa-tion.” Working Paper No. 61, University of ZurichInstitute for Empirical Research in Economics.

Berle, Adolf A. Jr. and Gardiner C. Means.1932. The Modern Corporation and Private Property.New York: Macmillan Company.

Bertrand, Marianne and Sendhil Mullai-nathan. 2001 “Are CEOs Rewarded for Luck?The Ones Without Principals Are.” Quarterly Jour-nal of Economics. August, 16:3, pp. 901–32.

Bettis, J. Carr, John M. Bizjak and Michael L.Lemmon. 2001. “Managerial Ownership, Incen-tive Contracting, and the Use of Zero-Cost Col-lars and Equity Swaps by Corporate Insiders.”Journal of Financial and Quantitative Analysis. Sep-tember, 36:3, pp. 345–70.

Bettis, J. Carr, Jeffrey L. Coles and Michael L.Lemmon. 2000. “Corporate Policies RestrictingTrading by Insiders.” Journal of Financial Econom-ics. August, 57:2, pp. 191–220.

Bizjack, John M., Michael L. Lemmon andLalitha Naveen. 2000. “Has the Use of PeerGroups Contributed to Higher Levels of Execu-tive Compensation?” Working Paper, PortlandState University.

Blanchard, Olivier Jean, Florencio Lopez-de-Silanes and Andrei Shleifer. 1994. “What doFirms do with Cash Windfalls?” Journal of Finan-cial Economics. 36:3, pp. 337–60.

Borokhovich, Kenneth A., Kelly R. Brunarskiand Robert Parrino. 1997. “CEO Contractingand Anti-Takeover Amendments.” Journal of Fi-nance. 52:4, pp. 1495–517.

Carpenter, Jennifer N. 1998. “The Exercise


and Valuation of Executive Stock Options.” Jour-nal of Financial Economics. 48:2, pp. 127–58.

Cheng, Shjun, Venky Nagar and Madhar V.Rajan. 2001. “Control versus Risk in Stock-BasedIncentives: Evidence from Antitakeover Regula-tion.” Working paper, University of MichiganBusiness School.

Core, John E. and David Larcker. 2002. “Per-formance Consequences of Mandatory Increasesin Executive Stock Ownership.” Working paper,Wharton School.

Core, John E., Wayne Guay and David F.Larcker. 2001. “Executive Equity Compensationand Incentives: A Survey.” Working paper, Whar-ton School.

Core, John E., Robert W. Holthausen andDavid E. Larcker. 1999. “Corporate Governance,Chief Executive Compensation, and Firm Per-formance.” Journal of Financial Economics. 51:3,pp. 371–406.

Crystal, Graef S. 1991. In Search of Excess. NewYork: Norton.

Cyert, Richard, Sok-Hyon Kang and PraveenKumar. 2002. “Corporate Governance, Take-overs, and Top-Management Compensation:Theory and Evidence.” Management Science. 48:4,pp. 453–69.

David, Parthiban, Rahul Kochar and EdwardLevitas. 1998. “The Effect of Institutional Inves-tors on the Level and Mix of CEO Compensa-tion.” Academy of Management Journal. 41:2, pp.200–08.

Fox, Justin. 2001. “The Amazing Stock OptionSleight of Hand.” Fortune. June 25, pp. 86–92.

Fried, Jesse M. 1998. “Reducing the Profitabil-ity of Corporate Insider Trading through Pre-trading Disclosure.” Southern California Law Re-view. 71:2, pp. 303–92.

Gillan, Stuart. 2001. “Has Pay for PerformanceGone Awry? Views from a Corporate Gover-nance Forum.” Research Dialogue. July, 68, pp.1–16 .

Habib, Michel A. and Alexander P.Ljungqvist. 2000. “Firm Value and ManagerialIncentives.” Working paper, London BusinessSchool.

Hall, Brian J. 1999. “A Better Way to PayCEOs?” in Executive Compensation and ShareholderValue: Theory and Evidence. Jennifer Carpenterand David Yermack, eds. Boston, Mass.: KluwerAcademic, pp. 35–46.

Hall, Brian J. and Kevin J. Murphy. 2000. “Op-timal Exercise Prices for Executive Stock Op-tions.” American Economic Association Proceedings.90, pp. 209–14.

Hallock, Kevin. 1997. “Reciprocally Interlock-ing Boards of Directors and Executive Compen-

sation.” Journal of Financial and Quantitative Anal-ysis. 32:3, pp. 331–34.

Hartzell, Jay C. and Laura T. Starks. 2002.“Institutional Investors and Executive Compen-sation.” Working paper, New York UniversityStern School of Business.

Jensen, Michael. 1986. “Agency Costs ofFree Cash Flow, Corporate Finance, and Take-overs.” American Economic Review. 76:2, pp.323–29.

Jensen, Michael and William Meckling.1976. “Theory of the Firm: Managerial Behav-ior, Agency Costs, and Ownership Struc-ture.” Journal of Financial Economics. 3:5, pp.305– 60.

Jensen, Michael and Kevin Murphy. 1990.“Performance Pay and Top Management Incen-tives.” Journal of Political Economy. 98:2, pp. 225–63.

Johnson, Marilyn F., Susan Porter and Marga-ret B. Shackell. 1997. “Stakeholder Pressure andthe Structure of Executive Compensation”Working paper, Michigan State University, Uni-versity of Massachusetts at Amherst, NotreDame.

King, Ralph. 2002. “Insider Loans: EveryoneWas Doing It.” Business 2.0. November; availableat �http://www.business2.com/articles/mag/0,1640,44304,00.html�.

Leonhardt, David. 2002. “It’s Called a Loan,But It’s Far Sweeter.” New York Times. February 3,Section 3, Page 1.

Levinsohn, Alan. 2001. “A Garden of StockOptions Helps Harvest Talent.” Strategic Finance.February, 82, pp. 81–82.

Lublin, Joan. 2002. “As Their CompaniesCrumbled, Some CEOs Got Big-Money Payouts.”Wall Street Journal. February 26, p. B1.

Murphy, Kevin J. 1999. “Executive Compensa-tion,” in Handbook of Labor Economics. OrleyAshenfelter and David Card, eds. Amsterdam:North Holland, pp. 2485–563.

Murphy, Kevin J. 2002. “Explaining Execu-tive Compensation: Managerial Power versusthe Perceived Cost of Stock Options.” Univer-sity of Chicago Law Review. Summer, 69, pp.847– 69.

Pearl Meyers and Partners. 2002. Executive PayTrends: Looking Forward and Looking Back.

Ofek, Eli and David Yermack. 2000. “TakingStock: Equity-Based Compensation and the Evo-lution of Managerial Ownership.” Journal of Fi-nance. 55:3, pp. 1367–384.

Rappaport, Alfred. 1999. “New Thinking onHow to Link Executive Pay with Performance.”Harvard Business Review. March, 77:2, pp. 91–101.


Seyhun, H. Nejat. 1998. Investment Intelligencefrom Insider Trading. Cambridge: MIT Press.

Shleifer, Andrei and Robert Vishny. 1986.“Large Shareholders and Corporate Control.”Journal of Political Economy. 94:3, pp. 461–88.

Shleifer, Andrei and Robert Vishny. 1989.“Management Entrenchment: The Case ofManager-Specific Investments.” Journal of Finan-cial Economics. 25:1, pp. 123–40.

Shleifer, Andrei and Robert Vishny. 1997. “ASurvey of Corporate Governance.” Journal of Fi-nance. 52:2, pp. 737–83.

Thomas, Randall S. and Kenneth J. Martin.1999. “The Effect of Shareholder Proposals on

Executive Compensation.” University of Cincin-nati Law Review. 67:4, pp. 1021–081.

Wade, James B., Joseph Porac and TimothyPollock. 1997. “Worth, Words, and the Justifica-tion of Executive Pay.” Journal of OrganizationalBehavior. 18:S1, pp. 641–64.

Williamson, Oliver. 1964. The Economics of Dis-cretionary Behavior: Managerial Objectives in a The-ory of the Firm. Englewood Cliffs, N.J.: PrenticeHall.

Yermack, David. 1997. “Good Timing: CEOStock Option Awards and Company News An-nouncements.” Journal of Finance. June, 52, pp.449–76.


Executive Compensation as an Agency Problem

Documents

Transcript of Executive Compensation as an Agency Problem